Methods, systems and computer readable media for quantifying and removing offset bias signals in a chemical array data set having one or more channels. In one embodiment, for each channel of data in the data set, a first set of features is selected from the data set. Surface intensities are calculated for features in the first selected set of features and surface intensifies of features not in the first selected set are calculated from the calculated surface intensities. A second set of features is selected, the intensity values of which are within a range of correspondingly located surface intensity values defined by upper and lower threshold intensities. Secondary surface intensifies are calculated for features in the second selected set of features and secondary surface intensities for all other locations on the array that were not locations corresponding to the features having secondary surface intensities calculated therefore, are calculated. Feature intensities of the channel features are then corrected as a function of the secondary surface intensities.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A method for quantifying and removing offset bias signals in a chemical array data set having one or more channels, said method comprising: for each channel of data in the chemical array data set, and using at least one computer processor, selecting a first set of features from the data set; calculating surface intensities for each feature in the first selected set of features; calculating a spread of the intensity values of the first selected set of features from the calculated surface intensities; selecting a second set of features the intensity values of which are within a range of intensity values between ± spread times a predetermined multiplier; calculating surface intensities for features in the second selected set of features; calculating surface intensities for all other features in the data set for the channel from the calculated surface intensities for each feature in the second selected set; correcting feature intensities of the channel features by subtracting the surface intensities for each feature in the second selected set at locations that correspond to respective locations of said feature intensities; and outputting the corrected feature intensities.
2. The method of claim 1 , wherein the selected features in said first set are negative control features.
3. The method of claim 1 , wherein said second set includes said first set.
4. The method of claim 1 , wherein surface intensities are calculated for all features in the second selected set.
5. The method of claim 1 , further comprising sampling the second selected set of features, wherein said calculating surface intensities is performed on a sampled, subset of said second selected set that is smaller than said second selected set.
6. The method of claim 5 , wherein said sampling is performed with a moving window filter.
7. The method of claim 1 , further comprising randomly selecting a predetermined percentage of features from the second selected set of features, wherein said calculating surface intensities is performed on the signal intensities of the randomly selected features, and wherein said predetermined percentage is less than one hundred percent.
8. The method of claim 1 , wherein at least one of said selecting steps further includes excluding features having non-uniform intensity distributions.
9. The method of claim 1 , wherein said calculating surface intensities for each feature in the first selected set of features comprises calculating a surface fit to intensity values of said features with a polynomial approximation algorithm.
10. The method of claim 9 , wherein said polynomial approximation algorithm is a second-order polynomial approximation algorithm.
11. The method of claim 1 , wherein said calculating surface intensities for each feature in the first selected set of features comprises locally-weighted, least-squares regression.
12. The method of claim 1 , wherein said calculating surface intensities for features in the second selected set of features comprises locally-weighted, least-squares regression.
13. The method of claim 1 , wherein said calculating surface intensities for features in the second selected set of features comprises calculating a surface fit to intensity values of said features with a polynomial approximation algorithm.
14. The method of claim 13 , wherein said polynomial approximation algorithm is a second-order polynomial approximation algorithm.
15. A method for quantifying and removing offset bias signals in a chemical array data set having one or more channels, said method comprising: for each channel of data in the chemical array data set, and using at least one computer processor, selecting a first set of features from the data set; calculating surface intensities for features in the first selected set of features and calculating surface intensities of features not in the first selected set, from said calculated surface intensities for features in the first selected set of features; selecting a second set of features the intensity values of which are within a range of correspondingly located surface intensity values defined by upper and lower threshold intensities; calculating secondary surface intensities for features in the second selected set of features and calculating secondary surface intensities for all other locations on the array that were not locations corresponding to the features having secondary surface intensities calculated therefore; correcting feature intensities of the channel features as a function of said secondary surface intensities; and outputting the corrected feature intensities.
16. The method of claim 15 , wherein said upper and lower threshold intensities are calculated as a function of said calculated surface intensities.
17. The method of claim 16 , wherein said function of said calculated surface intensities includes calculation of a spread of the intensity values of the selected set of features from the calculated surface intensities.
18. The method of claim 17 , wherein the spread is calculated by a robust technique, excluding outliers of the intensity values.
19. The method of claim 15 , wherein said correcting feature intensities comprises subtracting secondary surface intensities at locations that correspond to respective locations of said feature intensities.
20. A system for quantifying and removing offset bias signals in a chemical array data set having one or more channels, said system comprising: at least one computer processor; an interface through which one or more chemical array data sets are received by the system; and a program, stored in one or more memory components and executed by at least one computer processor that, for each channel, selects a first set of features from the data set; calculates surface intensities for each feature in the first selected set of features and calculates surface intensities of features not in the first selected set, from said calculated surface intensities for features in the first selected set of features; selects a second set of features the intensity values of which are within a range of correspondingly located surface intensity values defined by upper and lower threshold intensities; calculates secondary surface intensities for features in the second selected set of features and calculates secondary surface intensities for all other locations on the array that were not locations corresponding to the features having secondary surface intensities calculated therefore; and corrects feature intensities of the channel features as a function of said secondary surface intensities.
21. A non-transitory computer-readable media having instructions stored thereon that, when executed by a processor, cause the processor to perform the method of claim 1 .
22. A non-transitory computer-readable media having instructions stored thereon that, when executed by a processor, cause the processor to perform the method of claim 15 .
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
October 12, 2006
February 1, 2011
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.